Chain of trust

From ICMS
Revision as of 14:49, 2 February 2011 by Woozle (talk | contribs) (Created page with 'category:concepts category:commerce ==About== A chain of trust is a chain of personal connections between two individuals who do not necessarily know each other direc...')
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

About

A chain of trust is a chain of personal connections between two individuals who do not necessarily know each other directly. Each link in the chain involves some known amount of trust, such that the people at each end may rationally evaluate how much they trust each other without prior acquaintance.

chain of accountability

Part of the reason this can work is that a chain of trust carries with it a "chain of accountability".

A simplified example:

  • Elaine and Alice want to do business with each other, but they don't know each other personally
  • Bob trusts Alice
  • Charlie trusts Bob
  • Doris trusts Charlie
  • Elaine trusts Doris

Based on this chain of trust, Elaine could decide to extend Alice a certain amount of credit in order to make the transaction possible. If, however, Alice turns out to be less trustworthy than Bob claimed, each member of the chain will want to adjust their trust-values accordingly:

  • Bob will downrate Alice's reliability significantly
  • Charlie may downrate Bob's credibility
  • Doris may donwrate Charlie a little
  • Elaine might downrate Doris a little

Each of these adjustments makes it less likely that Alice will be extended credit beyond her means, the next time there is an opportunity. Although the chain of trust failed to make an accurate determination, it should be able to guess more accurately next time.

It's not yet clear what the best adjustment algorithm is, e.g. should all trust-links in the chain be downrated equally, or should the downrating be heaviest closest to the failure? Allowance also needs to be made for personal judgement of where the blame is concentrated, because there are probably too many factors to measure accurately -- but we should get some idea of what works best under simplified/ideal situations.